A patch-dynamic metacommunity perspective on the persistence of mutualistic and antagonistic bipartite networks

The structure of interactions between species within a community plays a key role in maintaining biodiversity. Previous studies found that the effects of these structures might vary substantially depending on interaction type, for example, a highly connected and nested architecture stabilizes mutualistic communities, while the stability of antagonistic communities is enhanced in modular and weakly connected structures. Here we show that, when network dynamics are modeled using a patch-dynamic metacommunity framework, the qualitative differences between antagonistic and mutualistic systems disappear, with nestedness and modularity interacting to promote metacommunity persistence. However, the interactive effects are significantly weaker in antagonistic metacommunities. Our model also predicts an increase in connectance, nestedness, and modularity over time in both types of interaction, except in antagonistic networks, where nestedness declines. At steady state, we find a strong negative correlation between nestedness and modularity in both mutualistic and antagonistic metacommunities. These predictions are consistent with the structural trends found in a large data set of real-world antagonistic and mutualistic communities.